PROJECT SUMMARY/ABSTRACT For the 3 million patients with breast cancer who have survived the initial treatment phase, routine surveillance for recurrence does not improve outcomes. Although it is possible, using imaging or blood tests, to find cancer recurrence before it produces symptoms or signs, large randomized controlled trials and Cochrane database systematic reviews have found that such surveillance offers no benefit in terms of survival or quality of life for patients with breast cancer. Testing, if unwarranted, involves unnecessary risks to the patient including radiation exposure, patient anxiety, and potential false positives resulting in unnecessary work-up that carries its own risks. Despite the evidence and guidelines advising against it, there is evidence that surveillance testing often occurs after initial treatment of breast cancer. Recent single institution and group model HMO studies and an older national study found evidence of that advanced imaging is used in about half of patients with early stage breast cancers. There is little information about surveillance using blood tests for tumor markers. Our long-term goal is to develop an improved understanding of drivers of surveillance test use and its impact on outcomes and to identify potential interventions to steer provider and patient behavior to improved outcomes and reduced cost (financial and radiation risk). Once the significant drivers of surveillance test use are identified, this information can be used to improve value of care in at least three ways. First, patients will have a better understanding of the potential risks?in terms of cost and radiation exposure?associated with surveillance testing. Second, since the US spends over $4 billion per year on imaging and other tests related to cancer care, even a small reduction in the utilization of expensive unnecessary surveillance tests would represent significant financial savings without increasing mortality. Third, if providers understand the effect, or lack of effect, of modern methods of surveillance testing on patient outcomes, including financial burden, radiation risk, and mortality, their willingness to adhere to guidelines, and patient outcomes, could improve. In this project we will create estimates of the variation in surveillance testing use nationally, and will generate the first data regarding the association between provider characteristics and practice patterns and surveillance testing rates (Aims 1 and 3). We will provide the first data on the impact of surveillance on patient and societal outcomes, for the first time estimating the costs of and radiation exposure due to testing done that is not consistent with guidelines (Aim 2 and 4). Finally, we will create the first qualitative data to allow us to generate hypotheses about why surveillance occurs, which could be used to develop interventions to reduce unnecessary testing that can be tested in future randomized trials.